Skip to main content

Full text of "The fauna of Rancho La Brea"

See other formats


qQE 

747 

C2M56 

1911 

v. 2 

VPAL 







f&AnXA 



C. Lewis Genda 

17. S. NotionaJ Museum 

Washington 25, D. C 



MEMOIRS 

OF THE 

UNIVERSITY OF CALIFORNIA 
Vol. 1. No. 2 



THE FAUNA OF RANCHO LA BREA 



PART II. CAN I DAE 



BY 

JOHN C. MERRIAM 



BERKELEY 

UNIVERSITY OF CALIFORNIA PRESS 
1912 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



Volume 1 

1. — Triassic Ichthypsauria, with Special Reference to the American Forms, by 

John C. Merriam. Pages 1-196, plates 1-18. September, 1908 . Price, $3.00 

2.— The Fauna of Rancho La Brea; Part I: Occurrence, by John C. Merriam. 

Pages 197-213, plates 19-23. November, 1911 . . . . .30 

The Fauna of Rancho La Brea; Part II: Canidae, by John C. Merriam. 
Pages 215-272, plates 24-28. October, 1912. .... .80 

Volume 2 

The Silva of California, by Willis L. Jepson. Pages 1-480, plates 1-85, maps 1-3. 

December, 1910. Price, in paper covers, $9.00; by express, $9.80. Price, bound 
in green buckram, $10.00 ; by express, $10.90. 

European orders for the Silva of California should be addressed to 
T. Fisher Unwin, Adelphi Terrace, London 



1 ml 

H// 

V, 2 
l/PAL 



MEMOIRS 



OF THE 



UNIVERSITY OF CALIFORNIA 



Vol. 1. No. 2 



THE FAUNA OF RANGHO LA BREA 



PART II. CANIDAE 



BY 

JOHN G. MERRIAM 



BERKELEY 
UNIVERSITY OF CALIFORNIA PRESS 
1912 



INTRODUCTION 






Remains of eanid forms constitute a large part of the total quantity of fossil 
remains found in the deposits at Rancho La Brea. Thus far six species have 
been recognized in the fauna at this locality: these species are, Canis dims 
Leidy ; Canis occidentalis furlongi Merriam, J. C. ; Canis milleri Merriam, J. C. ; 
Canis ochropus orcutti Merriam, J. C. ; Canis andersoni Merriam, J. C. ; and 
TJrocyon calif ornicus Mearns. Of these forms the great wolf, Canis dims, 
makes up more than half of the total number of specimens obtained. C. o. 
orcutti, though much rarer than C. dims, is known by a considerable number 
of specimens. The other species are quite rare. 

It is probably true that the species known from the deposits at Rancho La 
Brea varied to some extent in degree of susceptibility to entanglement in the 
tar, and therefore that the specimens found represent slightly different percent- 
ages of the number of individuals in existence during the period of deposition. 
It is also probable that some of the forms known were less characteristic of this 
region than of adjacent territory. There seems, however, little doubt that the 
proportion of remains of the several species obtained in the asphalt gives a 
fairly satisfactory approximation of the canid fauna of this region while the 
beds were accumulating. Though foreshadowing the fauna of the present 
period in the presence of a gray fox, a coyote, and a timber wolf, the difference 
between the canid life of this period and that of the present is emphasized in 
the dominance of the wolves of the Canis dirus type, in the presence of the 
peculiar short-faced Canis andersoni, and in the specific or subspecific difference 
in the coyote and the timber wolf. The gray fox, the only canid of the Rancho 
La Brea fauna which does not seem to differ distinctly from existing species, 
is known by very few specimens. 



[217] 



CANIS DIRTTS Leidy 

Plates 24 to 28 ; text figures 1 to 26 

Canis primaevus Leidy. Proc. Acad. Nat. Sc. Philad., 1854, p. 200. Jour. Acad. Nat. Sc., 
Philad.. 1856, ser. 2, vol. 3, p. 167. pi. 17, fig's. 11 and 12. (Not C. primaevus Hodgson, 1833.) 

Canis dims Leidy. Proc. Acad. Nat. Sc., Philad., 1858. p. 21. 

Canis indianensis Leidy. Jour. Acad. Nat. Sc, Philad., 1869, p. 36S. 

Canis indianensis Leidy. Geol. Surv. Terrs, vol. 1, Fossil Vertebrates, 1873. p. 230, pi. 31, 
fig. 2. 

Canis mississippiensis Allen. Am. Jour. Sc, ser. 3. vol. 11. p. 49, 1876. 

Canis lupus. Cope and Wortman. Indiana Geol. and Nat. Hist,. 14th Ann. Rep., part 2, p. 9. 
1884. 

Canis indianensis. Cope, E. D., Jour. Acad. Nat. Sc, Philad., 1895, ser. 2, vol. 9. p. 453. pi. 
21. figs. 14 to 16. 

Cards indianensis. Merriam, J. C. Univ. Calif. Publ. Bull. Dept. Geol., vol. 3, p. 288, 1903. 

Canis indianensis. Freudenberg, W., Geol. u. Palae. Abh.. N. F.. Bd. 9, Heft 3, S. 22, 1910. 

Canis indianensis. Merriam, J. C. Mem. Univ. Calif., vol. 1, no. 2, p. 213, 1911. 

Type specimen, an tipper jaw with the cheek teeth, now in the collection of 
the Academy of Sciences, Philadelphia. Type obtained from Pleistocene beds 
on the Ohio River near Evansville, Indiana. 

The forms referred to this species include some of the most remarkable 
known representatives of the Canidae, and deserve particular mention with 
reference to structure, geological occurrence, geographic range, and taxonomic 
position. Up to the time of discovery of the deposits at Rancho La Brea this 
wolf was known only by very fragmentary remains, and the literature contains 
no adequate description of any phase of its structure. 

Canis dims is the most abundant and most important of the canids from 
Rancho La Brea, and was evidently the dominant type of wolf in this region 
at the time of deposition of the asphalt beds. This species includes the largest 
individuals of the Canis group known from America. Some of the specimens 
exceed in dimensions all the largest known Recent wolves. Other individuals 
are considerably smaller than some of the large northern wolves of the present 
day. The skull is especially large, and the head seems to have been relatively 
large compared with the limbs. The teeth are very massive, but those regions 
of the cheek-tooth dentition constructed especially for crushing are relatively 
small. 

The comparatively light limbs and very massive head show that the animal 
was not as well developed for running as the timber wolves and coyotes. The 

[ 21S ] 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 219 

massiveness of the dentition without corresponding development of the crushing 
surface indicates use of the teeth in smashing large bones. The form of the 
skull suggests that the head was normally held low and was often used in hard 
pulling and hauling of heavy bodies. The great number of individuals of 
G. dirus found at Rancho La Brea suggests that the wolves of this species some- 
times associated themselves in packs, and that groups of considerable size may 
have assembled to kill isolated ungulates and edentates. Particularly the young, 
aged, and injured, when they could be separated from their associates, would 
be the natural prey of the great wolf, but adults in normal strength may also 
have succumbed to the combined attack of several of these powerful annuals. 

History of Literature on Canis dirus 

In 1854 Dr. Joseph Leidy 1 described from deposits occurring on the banks 
of the Ohio River, a short distance below Evansville, Indiana, a collection of 
fossil bones including the remains of Megalonyx jeffersonii, Tapirus liaysii, 
Equus americanus, Bison americanus, Gervus virginianus, and a large wolf. 
The wolf remains consisted of an almost complete left maxillary containing all 
but one of the cheek teeth. The species represented seemed to Leidy to vary 
far enough from any existing form to require a distinct specific designation, and 
was accordingly described as Ganis primaevus. Leidy realized that others 
might fail to recognize the species as distinct, as is indicated in the following 
statement taken from his paper : ' ' Certain naturalists may regard the fossil as 
an indication of a variety only of Ganis lupus., and of the correctness of this view 
I will not attempt to decide." In 1856 Leidy 2 figured and redescribed the 
species under the same name. 

In a description of Ganis (Aelurodon) saevus, published in 1858, Leidy 
refers to the wolf previously described by him as Ganis primaevus, as follows : 3 

"The present extinct species is not so large as the one whose remains have been discovered 
in association with those of Megalonyx, Tapirus, Equus, etc., on the banks of the Ohio River. 
Indiana, to which the name of Canis primaevus was inadvertently applied (Proc. Acad. Nat. 
Sc. VII, 200; Jour. Acad. Nat. Sc. iii, 167), and which may now be distinguished by that of 
Canis dirus." 

In 1869 Leidy 4 again referred to the original specimen, which he had 
described as Ganis primaevus, and gave to it the name Ganis indianensis. 
Apparently Leidy had forgotten the reference to this form under the name of 
Canis dirus in 1858. The name Ganis indianensis has come to be the designation 
for this form co mm only used in the literature. The writer is indebted to Dr. 
O. P. Hay for the discovery that Leidy's use of Ganis dirus for this form pre- 
ceded its designation as Ganis indianensis. 

i Leidy, J., Proc. Acad. Nat. Sc. Philad., vol. 7, p. 200, 1854. 

2 Leidy, J., Jour. Acad. Nat. Sc. Philad., ser. 2, vol. 3, p. 167, pi. 17, figs. 11 and 12. 1S56. 

3 Leidy, J., Proc. Acad. Nat. Sc. Philad., p. 21, 1S58. 
* Leidy, X, Jour. Acad. Nat. Sc. Philad., p. 368, 1869. 



220 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 

In 1884 E. D. Cope and J. L. Wortman, in describing the post-Pliocene verte- 
brates of Indiana, 5 reviewed Leidy's description, and after careful considera- 
tion of the measurements of the type specimen concluded that it would be 
impossible to admit this fossil to the rank of a distinct and well-defined species, 
but it appeared, in their judgment, to be but a variety which has a representative 
in the mountains of Oregon today. 

In 1895 Cope, 6 in describing a specimen representing a large fossil wolf 
from Texas, referred to Leidy's type specimen as representing a distinct species, 
Canis indianensis. 

Previous to the discovery of Canis dims at Rancho La Brea, excepting the 
type specimen, the only materials described which had been referred to this 
form consisted of two specimens from California and one from Texas. Several 
limb bones described from the lead region of the Upper Mississippi by Allen 7 
represent a form evidently nearer to C. dims than to any other American 
species, and not separated from it by any characters mentioned in the original 
description. 

Tbe California specimens first referred to C. indianensis consisted of a lower 
jaw which Dr. Lorenzo Yates obtained from a Quaternary deposit in Livermore 
Valley. The Yates specimen (see fig. 25) was tentatively, referred by Leidy s 
to this species. 

In 1903 a fragment of a lower jaw with the canine, the sectorial, and the 
last premolar, obtained from an asphalt deposit in Tulare Coimty, California, 
was referred by Merriam 9 to C. indianensis. 

A fragment of a lower jaw referred to C. indianensis in Sinclair's report on 
Potter Creek Cave, 10 on a determination by Merriam, possibly represents a 
timber wolf or another large wolf closely related to this species. 11 

The Texas specimen referred to C. indianensis consisted of portions of 
an upper dentition, including M 1 , P 3 , the canine, and an incisor. It was obtained 
in the Equus horizon of the Tule Canon, on Staked Plains of Texas Iry W. F. 
Cummins. Cope, 12 to whom the specimen was submitted, pointed out some differ- 
ences between the teeth of this animal and those of Leidy's type, but was 
inclined to regard it as an individual of the same species. 

So far as is known to the writer, the first mention of the occurrence of Canis 
dims in the deposits of Rancho La Brea appears in the preliminary description 
of this localitv bv Merriam 13 in 1906. 



^ Cope, E. D., and Wortman. J. L., Indiana Geol. and Nat. Hist., 14th Ann. Rep., part 2. p. 9, 1S84. 

o Cope, E. D.. Jour. Aead. Sc. Philad., ser. 2, vol. 9, p. 453, 1895. 

" Allen, J. A., Am. Jour. Sc, ser. 3, vol. 11, p. 49, 1876. 

8 Leidy, J., Proe. Acad. Sc. Philad., p. 260, 1873; and Geol. Surv. of Terrs., vol. 1, Foss. Verts., p. 230, 1873. 

a Merriam, J. C, Univ. Calif. Publ. Bull. Dept. Geol., vol. 3, p. 288, 1903. 

io Sinclair, W. J., Univ. Calif. Publ. Amer. Arch. Ethn., vol. 2. p. 17, 1904. 

n See page 245 of this paper. 

i= Cope. E. D., Jour. Acad. Sc. Philad.. ser. 2, vol. 9, p. 454, 1895. 

is Merriam, J. C, Science, n.s., vol. 24, pp. 24S-250, Aug., 1906. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 221 

Figures of a nearly complete skeleton and of a perfect skull of this species 
were used in illustration of popular articles published by Merriam in 1908 and 
1909. 14 

The most recent discovery of remains of Canis dirus is that reported by 
Freudenberg 15 from Tequixquiac, Mexico. At this locality the posterior region 
of a skull has been found which closely resembles the specimens from Rancho 
La Brea. 

Geological Occurrence and Geographical Distribution of Canis dirus 

The fairly authenticated occurrences of Canis dirus known to the writer 
include Leidy 's type specimen from Indiana ; the material from the lead region 
of the Upper Mississippi described by Allen; a specimen from the Sheridan 
formation of Kansas, now in the American Museum; Cope's specimen from 
Texas; the California specimens from Livermore Valley, Tulare County, and 
Rancho La Brea; and the Mexican specimens described by Freudenberg. All 
of the material referred to G. dirus has been obtained from deposits held to be 
of Pleistocene age. There is reason for believing that the horizons at which 
these specimens have been found at different localities do not differ greatly, but 
a discussion of the time relation of these occurrences is best considered in a 
division of this memoir following a discussion of the fauna. 

The collection of remains associated with the type specimen, including, as 
it does, on the one hand, such extinct forms as Megalonyx jeffersonii, Tapirus 
hagsii, and Equus americanus, and, on the other hand, the Recent Bison ameri- 
canus and Gervus virginianus, cannot represent other than Pleistocene time. 

The Texas specimen described by Cope seems to have occurred at the same 
horizon as Mylodonf sodalis, Elephas primigenius, Equus excelsus, E. semi- 
plicaius, E. tau, E. major, Holomeniscus sulcatus, and H. macrocephalus. This 
assemblage including the three genera Equus, Elephas and Holomeniscus must 
be considered as Pleistocene. 

The lower jaw which Leidy described from Livermore Valley, California, 
is presumed to have been asociated with a number of remains representing other 
mammalian forms. This material includes the type specimen of a very large 
cat, Felis imperialis, a bison referred to Bison latifrons, and a large camel. 
The presence of Bison latifrons may be considered as evidence of Pleistocene 
age, though it is uncertain how closely the specimens were associated. 

The fragment of a jaw from Tulare County described by Merriam was asso- 
ciated with a portion of the skull of an edentate nearly related to Mijlodon, and 
is presumably Pleistocene. 

The material from Rancho La Brea described in the following paper is 



14 Merriam, J. C, Sunset Magazine, Oct., 1908, p. 172. Also Harper's Weekly, Dec. IS, 1909, p. 11. 
is Freudenberg, W., Geol. u. Palae. Abh., N.F., Bd. 9, Heft 3, p. 22, 1910. 



222 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 

associated with a fauna which falls well within the limits of the Pleistocene, 
A discussion of the definite stage of the Pleistocene represented is deferred to 
a later chapter, in which the evidence from various sources will be assembled. 

The material from the Upper Mississippi region is held to be Pleistocene, 
and the Sheridan formation in which the Kansas specimen was found is gener- 
ally considered to represent an early phase of the Pleistocene. 

The material described from Mexico by Freuclenberg seems to have been 
derived from the same horizon as the great lion, Fells atrox, as is the case at 
Eancho La Brea, and presumably represents a horizon near that of the Pleis- 
tocene of Rancho La Brea. 

The known occurrences of Cams dims show that this animal certainly roamed 
over a large part of the Mississippi Valley ; its range extended south into Mexico, 
and west into middle and southern California. Until we have a more exact 
determination of the time relations of the beds in which this species is found, 
it is not possible to be certain as to contemporaneity of the occurrence in 
all of these regions, but such evidence as is before us indicates that the forma- 
tions concerned do not differ greatly in age. It is probable that the species was 
at one thne present in all of the regions mentioned, though the earliest and 
latest occurrences may have differed much in the several regions. 

That the range of ('cutis dims extended considerably beyond the territory 
marked out by known occurrences is probable, but it is by no means certain 
that it covered a region as large as that now occupied by wolves of the C. occi- 
deittalis group. Until we are better acquainted with the correlation problem of 
the American Pleistocene, it is perhaps unsafe to attach much significance to 
the possible absence of C. dints from the Pleistocene of Silver Lake, Conard 
Fissure, Samwel Cave, and Port Kennedy Fissure, and its absence or rarity in 
Potter Creek Cave. Absence from some of these faunas may be due to differ- 
ence in age of the beds, but the deposits included in this list represent a wide 
range of the Pleistocene, and it is not probable that all are so far removed in 
thne from the beds containing C. dims as to have missed completely the life 
range of that species. Some of the localities, particularly the cave regions, 
evidently constituted a habitat very different from that of the known occur- 
rences of C. dims, and to this difference in environment the presence or absence 
of the great wolf may be due in some measure. The faunas of Potter Creek 
Cave and Samwel Cave in California lived in a hilly or mountainous country 
covered to a large extent with forest, whereas the Rancho La Brea fauna 
represents the life of a plain bordering the hills. In view of what is known, 
the great wolf may be presumed to represent a fauna which ranged mainly 
over the great plains of an area corresponding approximately to what is now 
the Sonoran region. What we know of the structure and probable habits of 
C. dims would be in agreement with such a range, as the animal seems par- 
ticularly suited for preying upon some of the larger plains mammals. 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 223 

Diagnostic Charactees 

The largest species of Canis known from the faunas of North America. Form 
and proportions in general near those of the existing timber wolves; head 
relatively larger and feet relatively smaller than in the large Recent timber 
wolves of the Canis pambasileus type. Skull attaining a length of 310 mm. 
or more; relatively broad across the palate, frontal region, and zygomatic 
arches. Sagittal crest high, inion showing an extraordinary backward projec- 
tion. Posterior extremities of nasal bones extending relatively far back. Nasal 
processes of frontals relatively short. Postpalatine foramina opposite posterior 
ends of superior carnassials. Optic foramen and anterior lacerated foramen 
close together in a common pit. Upper and lower carnassials relatively large 
and massive. P* with reduced cleuterocone, M 1 with greatly reduced hypocone, 
Mi with a small metaconid, P 2 and P 2 often without posterior cusps or tubercles. 



Skull 

The skull in this species (figs. 1, 2, 3, and 4) is larger than the cranium of any 
other wolf known to the writer. The basal length from the anterior end of 
the premaxillaries to the posterior side of the occipital condyles in one of the 
large specimens measured (no. 10856) is 282 mm. The total length of the 
skull projected on the plane of the palate, including the extraordinary back- 
ward projection of the inion, may be more than 310 mm. In spite of the great 
length, the width measured across the palate, between the orbits, across the 
postorbital processes of the frontals, and across the zygomatic arches, is 
relatively large, making the skull very massive. 

The facial region is characterized by extraordinary backward extension of 
the nasal bones, which reach a short distance behind a line connecting the most 
nearly approaching points on the orbits. The nasal processes of the frontals 
are generally relatively short. In the lateral region of the face an area at the 
anterior root of the zygomatic arch is generally very sharply depressed and 
forms a characteristic feature. On the inferior side of the zygomatic arch the 
attachment of the masseter muscles is very strongly marked, and ends anter- 
iorly on a prominent knob situated at the most inferior point on the suture 
between the jugal and the maxillary. 

The frontal region is relatively broad, the postorbital processes of the 
frontals being very largely developed. The frontal region shows only a slight 
median depression, which may be somewhat accentuated toward the nasal and 
parietal borders of the frontals. The sagittal crest is uniformly high, and is 
characterized especially by the extraordinary backward extension, which much 
exceeds that of the other wolves. 



224 



MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 





Figs. 1 and 2. Canis dirus Leidy. Skull, no. 10S34, X %. Fig. 1, lateral view; fig. 2, superior view. Eancho 
La Brea Beds. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 



22.3 




3 




._ , ~* 

4 

Figs. 3 and 4. Cams dims Leidy. Skull, no. 10834, X %. Fig. 3, posterior view; fig. 4, inferior view. Raneho 
La Brea Beds. 

Percentages of width of the skull at various points in relation to basal length measured from 
the anterior end of the premaxillaries to the posterior side of the occipital condyles are as follows : 

C. dirus(a) C. pambasileus(Z)) C. latrans 
Width across palate, measured between outer sides of 

superior sectorials 38.1% 34 % 32 % 

Width between most nearly approaching points on 

upper border of orbits 22.7% 20.S% 16.5% 

Width between postorbital processes of frontals 33.3% 27.2% 24 % 

Width across zygomatic arches 62 % 55.6% 59.4% 

(a) No. 10856. All numbers, unless otherwise noted, are from the series of the University of California 
Collections in Vertebrate Palaeontology. 

(6) No. 984, Univ. Calif. Mus. Vert. Zool. From Susitna River, Mt. MeKinley region. Alaska. This is 
according to Elliot (Field Columb. Mus. Publ., Zool. ser., vol. 6, p. 374) the largest Recent species of the 
North American wolves. 



226 MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 

The palate is relatively broad, and the posterior palatine foramina are set 
relatively far back. The posterior end of the vomer extending backward between 
the vertical plates of the palatines commonly reaches very slightly beyond the 
posterior nasal opening. This is in decided contrast to the Recent wolves, in 
which the broad posterior end of the vomer reaches well beyond the posterior 
nasal opening. In most of the specimens of Canis dims the posterior nasal 
opening is relatively wider at the anterior end than in wolves of the C. occi- 
dentalis type. 

The occipito-sphenoidal region of the skull exhibits a tendency to shortening, 
the distance from the posterior border of the glenoid fossa to the posterior 
side of the occipital condyles averaging slightly less than in other forms. On 
the basal occipital the uneven surfaces corresponding to the attachment of the 
longus capitis muscle are marked by very rough areas which commonly do 
not extend as far forward as in the timber wolves. In other forms these areas 
extend forward for a considerable distance anterior to the tympanic bullae. 
In the space between the tympanic bullae and the anterior end of the inferior 
notch of the foramen magnum the median ridge extending from the basal 
occipital to the basal sphenoid is usually relatively prominent and acute, while 
in the timber wolves it is commonly a low, broad, horizontally truncated ridge. 

The posterior aspect of the skull presents in general a very different form 
from that of the Recent wolves. The two ridges which form the lamboidal 
crest tend to converge sharply above the occiput, while in most Recent forms they 
sweep outward rather widely before uniting at the inion. On the other hand, 
the short processes formed at the lower ends of the transverse ridges and 
immediately behind the superior side of the posterior root of the zygomatic 
arches average smaller than in the modern species. In the superior region of 
the occiput the two lateral depressions in which the attachments of the rectus 
capitis posticus are situated average extraordinarily deep, and are separated 
by a narrow ridge which is relatively prominent with reference to the occiput 
as a whole. The backward projection or overhanging of the inion is extraor- 
dinary, as are also the height and thinness of the sagittal crest rising above it in 
posterior view. 

The lower jaw tends to be relatively longer than in most wolves and is also 
relatively higher and thicker below the inferior carnassial. Although the coro- 
noid process is not unusually high, the masseteric fossa is generally very deep 
and rough, indicating an unusually strong muscular attachment. 

Although the foramina of the skull do not in general vary greatly in form or 
position from those of the Recent species, certain minor modifications are 
usually noticeable. The posterior palatine foramina are commonly situated 
farther back than in the Recent American species, and are nearly opposite 
the posterior borders of the superior sectorials. The openings of the optic 
foramen and the anterior lacerated foramen are situated very near together in 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 227 

an uncommonly deep depression. The foramen ovale and the posterior opening 
of the alisphenoid canal are also situated very near together, and are usually 
in a distinct common depression. The situation of these foramina in this species 
is approximated in some of the Recent wolf species, but the openings are not 
often so nearly united in the Recent forms as to approach the extreme of 
variation here. 

Measurements of Skull 



m co ^ r 

3 S - — 

.a a - — 

■3 S 3 a 

'g "3 'S £ 'a 

tS S n r. - 

O O ^ — ^- 

Length, anterior end of premaxillaries to posterior end 

of occipital condyles 282 mm. 267 250 223 

Length, anterior end of premaxillaries to anterior end 

of posterior nasal openings 155 141 131.5 119 

Width across nose, measured between outer sides of 

bases of canines 67.3 58.5 52 50.2 

Width, measured between outer sides superior sec- 
torials 107.5 96.2 S5 89.5 

Width across zygomatic arches 175 ap 164.5 138.4 134 

Least diameter between superior borders of orbits 64.9 54.1 51 43.2 

Width between postorbital processes of frontals 93.9 77 67.7 64 

Length from a line drawn between posterior borders of 
glenoid fossae to posterior end of occipital con- 
dyles _... 54 57 52 52 

Length, anterior end of left ramus of mandible to 

middle of condyle 230 ap 210.5 202 17S 

Height of lower jaw measured between summit of cor- 

onoid process and inferior side of angle 91.3 87 85.7 82.5 

Height of lower jaw below hypoconid of M t 39.7 35.3 30.6 29.8 

Height of lower jaw below protoconid of P s 36.9 32.5 31.9 24.5 

Thickness of lower jaw below protoconid of M t 20.3 19.3 14.9 16 

(a) No. 10856. (c) No. 984, Univ. Calif. Mus. Vert.. Zool. 

(6) No. 10834. (d) No. 11257. 

ap approximate. 

Permanent Dentition 

Superior Dentition. — The incisors of the upper jaw (fig. 4) are not more 
noticeably crowded in specimens 10828 and 10856 than in the modern Alaskan 
wolves. 

I 1 is not materially different from that in the large modern wolves. On F 
the basal lobe on the median side is small or wanting and there is a second 



228 MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 

minute basal tubercle on the lateral side. I 3 is not materially different from 
the corresponding tooth in the modern wolves. 

The superior canine seems to be relatively short in anteroposterior diameter 
at the upper margin of the enamel, its diameter slightly exceeding that of 
considerably smaller specimens of the Alaskan wolf. The transverse diameter 
or thickness is relatively large, and the tooth shows a more nearly circular 
cross-section. The anterointernal enamel ridge usually runs nearly straight 
up to the margin of the enamel, without turning backward as in Recent Alaskan 
wolves. 

P 1 is sometimes smaller in anteroposterior diameter than P 1 of the large 
Recent wolves. 

P" is relatively much smaller compared with the sectorial than in the large 
Alaskan wolves, the anteroposterior diameter being about one-half that of P\ 
This tooth is frequently almost simple-crowned, as is exemplified in no. 10856, 
in which there is only the merest trace of a tubercle on the posterior border of 
the cingulum. In other cases (no. 10893) there is a distinct posterior basal 
cusp, and behind it a posterior basal tubercle. 

P 3 is also relatively small compared with the sectorial. (Ratio of antero- 
posterior diameter in P 3 and P 4 19:32 in G. dirus no. 10856, and 16.5:24.5 in 
a C. pambasileus from Alaska). There is a distinct posterior cusp present, 
and a very small basal tubercle is usually developed on the posterior border 
of the cingulum. The small posterior basal tubercle seems to be present more 
frequently than in the Recent wolves. The protocone is in general not made 
relatively smaller than in the Recent species through the regular addition of 
the posterior basal tubercle. The protocone commonly tends to be relatively 
la rge anteroposteriorly . 

P 4 is relatively large compared with all of the other premolars. The blade 
is massive, though not much thicker transversely in relation to the size of the 
skull in general than in the Recent wolves. The deuterocone is usually small, 
and the root supporting it does not ordinarily project as far toward the median 
line as in the modern wolves. Though usually distinctly set off from the pro- 
tocone, the deuterocone is in some cases entirely reduced. 

A peculiar feature appearing in quite large percentage of the specimens is 
found in the development of a sharply angular ridge on the lower side of the 
cingulum on the outer side of the tritocone. The character of this ridge is 
in general similar to that of the lower side of the cingulum on the outer side 
of the upper molars. In one instance (no. 10830) this ridge is very largely 
developed on the left P\ and several distinct, rounded tubercles have arisen 
upon it. This last instance may possibly be attributed to pathological changes 
or to injury. It presents an interesting tendency of development, though it may 
have been stimulated by extraordinary conditions. 

M 1 is relatively small both anteroposteriorly and transversely compared 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 229 

with the carnassial. The most noticeable peculiarity of this tooth is the reduc- 
tion of the hypocone, which is smaller than in any of the Eecent North 
American wolves. The short ridge of the hypocone never extends forward 
around the base of the protocone to join an anterior basal ridge as commonly 
occurs in the Recent wolves. The extreme median side of the hypocone fre- 
quently does not extend farther toward the median line than does the base of 
the protocone. The protocone, paracone, and metacone are not materially dif- 
ferent from the corresponding tubercles of the modern wolves, though the bases 
of the paracone and metacone often tend to be relatively thick transversely. 
A distinct metaconule is always present and a small protoconule is usually 
developed. 

M" is relatively small compared with the sectorial, and is usually also small 
compared with M 1 . In specimen 10856 its dimensions are practically identical 
with those of a Recent C. panibasileus which has a much smaller skull, and 
in which the other teeth are smaller. The principal difference between the 
form of this tooth seen here and that in the Recent wolves is found in the 
uniformly smaller size of the hypocone. In some instances the anterior end 
of the hypocone crest does not extend forward as a ridge of the cingulum along 
the anterior side of the tooth. 

Inferior Dentition. — In most specimens observed the incisors are thrown 
considerably out of alignment by lateral crowding, I 2 being set back at least 
as far as L, while the posterior border of L, is nearly even with the anterior 
border of L. 

I 3 has much the same form as in the large Recent Alaskan wolves excepting 
that the lateral lobe tends to be relatively small. 




Fig. 5. Canis dims Leidy. Superior view of inferior premolar and molar series, no. 10S34, natural size. 
Rancho La Brea Beds. 

In the inferior canine the sharp anterointernal ridge which marks the 
enamel of this tooth takes a course somewhat different from that in the Alaskan 
C. panibasileus. In passing forward from the posteroinferior region of the 
inner face of the tooth it does not rise as rapidly as in the Recent Alaskan 
wolves, but extends forward to a point near the anterior side of the tooth and 
only a few millimeters above the base of the enamel before it takes a direct 
upward course toward the apex of the crown. The angle formed by the sharp 
upward turning of this ridge is higher up on the side of the tooth, and conse- 
quently more obtuse in the Recent Alaskan forms. In a specimen of C. latrans 
from Manitoba this enamel ridge is closely similar to that in C. dims. 

Pi (figs. 5, 6, and 7) tends to be rather small compared with P». A faint 



230 



MEMOIKS OF THE UNIVERSITY OF CALIFORNIA. 




indication of an anterior basal tubercle occurs less frequently and is less dis- 
tinct tban in tbe Alaskan wolves. A faint posterior basal tubercle is present 
in some instances. Ordinarily both anterior and posterior basal tubercles are 
absent. 

P 2 in most cases shows no trace of either anterior or posterior tubercles, the 
tooth being as simple in form as the anterior lower premolars of Temnocyon. 
In a few cases, as in specimen 10727 (fig. 7), a well-developed posterior basal 
cusp is present with a minute posterior basal tubercle situated behind it. 

P 3 possesses a distinct posterior cusp, and a minute posterior basal tubercle 
is sometimes present. The posterior shelf or heel of the cingulum may be prom- 
inent and rather sharply turned up on the posterior side, as in the modern 

wolves ; or may, as in no. 10856, be much less 
distinctly marked, and may slope downward 
from the posterior basal cusp to the poster- 
ior margin of the tooth without exhibiting 
any upward curvature. 

P4 is relatively large compared with P 3 , 
but shows almost exactly the same size in 
relation to Mi as is seen in the Alaskan C. 
pamiasileus. The posterior cusp and pos- 
terior basal tubercle are well developed. 
The posterior portion of the tooth is some- 
what wider than the anterior in many cases. 
The posterior basal tubercle is usually situ- 
ated on the extreme posterior portion of the 
cingulum, and its posterior border is the 
extreme posterior margin of the tooth. In some cases, as in no. 10727, this 
tubercle is separated from the posterior border by a distinct notch and another 
small tubercle is present on the posterior margin of the cingulum. 

In M 1 , the trigonid portion of the tooth is generally relatively long and mas- 
sive, or the heel region is relatively short compared with the large Recent wolves 
of North America. The metaeonid is also generally smaller than in the 
modern species, though not always smaller than in all the Recent varieties. 
The hypoconid portion of the heel exhibits a tendency toward relatively greater 
development than the entoconid region. In nearly all specimens three small 
tubercles have developed in the space between the metaeonid and the entoconid. 
One of these is usually situated on the base of the metaeonid, one on the base 
of the entoconid, and one intermediate between the two. In some cases the 
number of these secondary tubercles varies above or below three, and their 
position may also vary somewhat from the situations indicated as most typical. 
M 2 and Ms, the tubercular molars, are relatively small both anteroposterior!}' 




Figs. 6 and 7. Canis dints Leidy. Lateral 
view of inferior premolar series. Fig. 6, 
no. 10S34, natural size; fig. 7, no. 10727, 
natural size. Rancho La Brea Beds. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 



231 






il„ 



9 10 

Figs. 8 to 10. Cunis dirus Leidy. 

superior view, natural size. Fig. 
8, no. 19472; fig. 9, no. 19473; fig. 
10, no. 19474. Rancho La Brea 
Beds. 



and transversely in comparison with the inferior carnassial. In specimen 
10856 M 2 shows an anteroposterior diameter of 
12.8 mm., in comparison with 35.7 mm. in the car- 
nassial. In the large Alaskan wolf, G. pamliasileus, 
(no. 984, Univ. Calif. Mus. Vert. Zool.) the corre- 
sponding dimensions are M 2 , 12 mm.: carnassial 
29 mm. In Mr. the relative reduction is still more 
noticeable, this tooth measuring 7 mm. anteropos- 
teriorly in the G. pambasileus specimen and only 
6.5 mm. in G. dirus (no. 10856) . The reduction in M 2 is due in some cases to 
a noticeable weakness of the heel region. The metaconid of M 2 is in some 
cases reduced to a small tubercle situated on the side of the nearly central 
protoconid. In rare cases a small tubercle is present in the paraconid region 
(fig. 8). As a rule no tubercle is present to represent paraconid or parastvlid 
(fig. 10). 

A very large specimen (no. 11281) referred to G. dims presents a number 
of characters in which the dentition varies slightly from the typical individuals 
of this species. (See fig. 11). At the same time the principal peculiarities of 
this specimen are those of the C. dirus form. P 4 is exceptionally massive, the 
transverse diameter of the protocone blade 



being relatively large. Another 
upper carnassial with an anteroposterior 
diameter of 35.2 mm. (no. 12576) is relatively 
a little thinner transversely. In M 1 of speci- 
men 11281 the hypocone is larger than in any 
other individual found. It does not, however, 
reach the relatively large size of the Recent 
American wolves, nor is the hypocone ridge 
extended around the anterior side of the protocone as in the wolves of the C. 
occidentalis type. In M 2 also the hypocone is relatively a little larger than in 
the average of the Rancho La Brea individuals of G. dirus. The metacone is 
also relatively a little larger than in average specimens from this locality, 
having, as in Leidy 's type, nearly the size of the paracone. 

A lower carnassial associated with no. 11281 also indicates an individual of 
gigantic size. The characters of this tooth are those of the typical G. dirus. 

Specimen 11281 evidently belongs in the true C. dirus group, but represents 
an exceptionally large form varying from typical individuals in the nature of 
the hypocone of M 1 . 



even larger 




Fig. 11 Canis dirus Leidy Upper carnassial 
and upper molars, occlusal view. no. 
11281, natural size. Bancho La Brea 
Beds. 



232 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 



Measurements op Permanent Dentition 



03 

a ? 



n « 

S c« 

.9 2 § 

.£ "9 O » .Eh 

ra 13 a Bus 

. 2 <a >» . 
O £« HO q<J 






I 3 , greatest transverse diameter 

C, greatest anteroposterior diam- 
eter at lower edge of enamel 17.5 mm. 

P 1? greatest anteroposterior diam- 
eter 7.7 

P,, greatest anteroposterior diam- 
eter 15.4 

P 3 , greatest anteroposterior diam- 
eter 16.7 

P 4 , greatest anteroposterior diam- 
eter 20 

M lt greatest anteroposterior diam- 
eter 35.7 

Mj, greatest anteroposterior diam- 
eter of heel, on outer side 9.2 

M,. greatest transverse diameter 

of heel 13.5 

Mj, greatest transverse diameter 

of trigonid 14.3 

M 2 , greatest anteroposterior diam- 
eter 12.8 

M 2 , greatest anteroposterior diam- 
eter of heel, on outer side 4.5 

M„, greatest transverse diameter 10 

M 3 , greatest anteroposterior diam- 
eter 6.5 

I 1 , greatest transverse diameter 



7.4 
15.3 
15.S 
19.5 
34.5 

8.8 

13 

13.6 

13.3 

5 
9.3 

7.3 



7.6 
15.5 

7 
13.5 
15.5 
16.5 
29 

S.2 

11 

11.8 

12.3 

4.5 
10 



T3 d 2 <3 » m m 
OO OtBW OB 



d 

^« s 

2 o © 

OPh O 





6.3 








14.7 








15.5 




16 


16 


19.9 




18.2 


32 


33.7 


3S.6 


32 




8.2 


9.3 


7.8 


12.2 


12.6 


13.7 


12 




13.8 




13.5 


12.66 


12.5ap 




12.4 
4.6 


9 


10 




9.1 
5.9 



I 3 , greatest anteroposterior diam- 
eter 12c 

C, greatest anteroposterior diam- 
eter at upper edge of enamel 17 

P 1 , greatest anteroposterior diam- 
eter 10.2 

P 2 , greatest anteroposterior diam- 
eter 16 



10.5 



16.2 



9.4 S.5 

16.2 14.8 15 



(a) No. 10856. 

(b) No. 10S34. 

(c) No. 984, Univ. Calif. Mus. Vert. Zool. 

(d) No. 115995, U. S. Nat, Mus. 



(e) No. 10828. 
(/■) No. 11282. 
(g) No. 11257. 
ap approximate. 



MERRIAM: THE FAUNA OP RANCITO LA BREA. 



233 



c4 

a 2 

3 oq O 

H .„ -^ 

• rH O U 

'S ho a 

• £ =« 

Q ,3 Ph 

P 3 , greatest anteroposterior diam- 
eter 19 

P 3 , greatest transverse diameter 

P 4 , greatest anteroposterior diam- 
eter 32 

P 4 , greatest transverse diameter 

across denterocone 16.2 

P 4 , greatest transverse diameter 

across protocone 13 

M 1 , greatest anteroposterior diam- 
eter 20 

M', greatest transverse diameter.. 24 

M 1 , transverse diameter of proto- 
cone 13.6 

M 2 , greatest anteroposterior diam- 
eter 10 

M-, greatest transverse diameter.... 15.4 

ap approximate. 

* Without posterior enamel. 

} Without enamel. 



■en 



C. dirus (6] 
medium spe 
Raucho La 


=h en 

>-. . 

BO 


m 

~ a 


a ~ 


'S- — 

'S x 






■f. 


| 


18.1 


18.2 


16.5 


17.33 






18.5 




17.5 


7.9 


8 










8 




7.4 


30.7 


32 ap 


24.5 


27 








33.8 


28.2 


15 14.5 ap 


14 


14.5 








16.9 


15.8 


13 




11 










15.3 


12.8 


18.7 


18.5* 


17.5 








18 


20 


16.4 


23 


21.5 
12.6 


22.5 


23 






21.3 


25 


20.7 
12.7 


9.2 


10 


10 










10.8 


9 


14.4 


14.9$ 


14 










14.6 


12.9 



Milk Dentition 

The milk dentition is well shown in several specimens. In no. 10831 it 
presents the following characters: The superior temporary carnassial (figs. 12 
and 13) has well-developed cutting blades; the inner root is situated almost 






13 



14 



15 



Pigs. 12 to 15. Canis dirus Leidy. Milk dentition. Figs. 12 and 13, no. 19475, natural size; figs. 14 and 15. 
no. 19481, natural size. Fig. 12, upper milk carnassial, inner side; fig. 13, upper milk carnassial and tuber- 
cular milk molar, occlusal view; fig. 14, lower milk carnassial and preceding tooth, inner side; fig. 15, 
lower milk carnassial, occlusal view. Rancho La Brea Beds. 



directly above the apex of the protocone; there seems to have been no denter- 
ocone present upon the base of the inner root, but there is a minute tubercle 
on the cingulum a short distance in advance of the base of this root, and nearer 
the normal position of the deuterocone on the permanent carnassial. In the 
inferior milk carnassial (figs. 14 and 15) the cutting blades are well-developed 



234 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 



and the metaconid is small; on the heel the entoconid approaches the size of 
the hypoconid, and the hypoconulid is near the size of the entoconid. Dm 3 (fig. 
14) possesses a large posterior cusp. A minute basal tubercle may be present 
on the anterior side of this tooth. 

Measurements of Milk Dentition 

No. 10S31 

Dm,, greatest anteroposterior diameter 8.7 mm. 

Dm 4 , greatest anteroposterior diameter 15.5 

Dm 4 , greatest transverse diameter across heel 6.6 

Dm 3 , greatest anteroposterior diameter 14.6 

Dm 4 , greatest anteroposterior diameter 10.8 

Dm 4 , greatest transverse diameter 11.4 

Vertebrae 

Owing to the peculiar occurrence of remains at Rancho La Brea it is very 
difficult to obtain complete connected skeletons. The viscosity of the asphalt 
mass in which the bones are entombed has permitted the elements of each 
individual to move easily in many directions. It has therefore been found very 
difficult to obtain connected parts such as occasionally appear in most deposits 
of fossil remains. Particularly exceptional is it to find a vertebral column which 
is even approximately complete, and in which all of the elements can be recog- 
nized as belonging to one specimen and certainly distinct from numberless 
other skeletons closely packed in the matrix. 

For the reasons just given it is difficult to make more certain of the vertebral 
formula of Canis dims than to learn that the relations of the vertebrae 
where found in connected portions of the column, and as indicated by the 
fitting together of the abundant material available, show no indication of a 
variation in number of vertebrae from that of the modern timber wolves. 





Fig. 16. Canis dims Leidy. 
Fig. 17. Canis dims Leidy. 



16 ~ 17 

Atlas, inferior view, no. 10834, X %. Bancho La Brea Beds. 
Axis, lateral view, no. 10S34, X %. Bancho La Brea Beds. 



The atlas is rather variable in form, but resembles that of Canis lupus in 
most respects. It differs from that of C. lupus mainly in the more common 
tendency of the transverse processes to take on a triangular form. In C. lupus, 



MBRRIAM: THE FAUNA OP RANCHO LA BREA. 235 

and in the wolves generally, the transverse processes of the atlas are hroad 
transversely, and are formed in such a manner as to give each transverse 
process an approximately quadrate form. In G. dirus the lateral margins 
of the transverse processes are sometimes so truncated that the form tends 
toward the quadrate ; but in many cases the margin, beginning with the anterior 
notch for the vertebral artery, slopes outward and backward, so that each pro- 
cess tends toward a more nearly triangular form with the apex at the postero- 
lateral angle. This form of atlas is shown especially well in figure 16. 

Compared with the only specimens of G. occidentalis available for studv, the 
atlas of G. dims is commonly distinguished ; 1) by the much larger opening of 
the vertebrarterial canal, which perforates the transverse process almost normal 
to the blade instead of perforating it obliquely; 2) by the much shallower an- 
terior notch for the vertebral artery; 3) by the less marked tendency of the 
transverse processes to project behind the posterior ends of the facets for 
articulation with the axis. 

The axis (fig. 17) is near that of Ganis lupus in form. In lateral view it 
tends to be short and relatively high compared with most canids. Seen from 
below it is relatively wide. It is generally characterized by relatively short 
transverse processes, which often project little if any behind the posterior end 
of the centrum. A character which seems generally to appear in the axis of 
this form is the presence of a shallow notch on each side in the postero-lateral 
margin of the spine between its postero-superior angle and the small tuberosities 
just above the posterior zygapoplryses. This region of the margin is in most 
canids without a distinct emargination ; there is, however, such a notch in the 
axis of G. occidentalis. 

The cervicals from number three to number seven do not differ markedly 
from those of the modern timber wolves. In number three the low spinal 
ridge is clearly marked in all specimens examined. The postero-superior 
tubercles above the postzygapophyses vary considerably in size in different 
individuals. The anterior and posterior limbs of the transverse processes extend 
slightly beyond the ends of the centrum. On number five the laterally directed 
posterior tubercles on the transverse processes are well developed. In number 
six the anterior side of the blade of each transverse process is commonly rather 
deeply notched immediately anterior to the base of the process arising laterad 
of the vertebrarterial canal. The inferior side of number seven is marked 
by a distinct median ridge. 

The dorsals do not differ markedly from those of G. occidentalis. The first 
dorsal is usually characterized by a relatively great transverse diameter of the 
postero-inferior region of the spine, and by the tendency to development of 
a deep longitudinal groove on this region of the spine. Dorsals three to ten 
are distinguished from those of some of the other wolves by the absence of 
notches between the anterior zygapophyses and the anterior border of the trans- 



236 MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 

verse processes. The anterior border of the lamina in this region is usually 
only slightly concave. The small tubercles which commonly appear immediately 
above the rib-articulation on the transverse processes of the third to the tenth 
dorsals are commonly rather small compared with those of many canids. On 
the twelfth and thirteenth dorsals the metapophysial processes are not dis- 
tinctly separated from the zygapophyses by an antero-median notch as in some 
other canids, but the inner side of the metapophyses is nearly continuous with 
the prezygapophysial faces. 

The lumbars are closely similar to those of the modern timber wolves. They 
seem generally to show the same tendency of the inner side of the metapophyses 
to grade into the inner face of the prezygapophyses shown in the most pos- 
terior dorsals. 

The sacrum varies somewhat in form, being wider posteriorly in some indi- 
viduals than in others, this difference being possibly due to sex. It does not, 
however, seem to depart distinctly from the form seen in Canis lupus. The 
posterior extensions of the transverse processes of the last vertebra included 
in the sacrum commonly extend only a short distance behind the posterior 
articular face of the centrum, but vary somewhat in this respect. The spines 
of the posterior vertebrae included in the sacrum are higher than in a specimen 
of C. occidentalis available for comparison. 

It is not easy to make certain of the form of the tail and of the number of 
vertebrae included in it; so far as can be determined, the number of caadals 
did not vary distinctly from that in C. occidentalis. The size of the tail as 
indicated by the form of the posterior region of the sacrum was not very dif- 
ferent from that of the modern timber wolves. The individual caudal centra 
are not found to differ in form from those of the timber wolves. 

"No peculiar characters are noted in the elements of the rib-basket and 
sternum. 

EXTREMITIES AND ARCHES 

The form and proportions of the limbs in Can is dims show that it was an 
animal constructed on much the same lines as the timber wolves, though some- 
what heavier, with the limbs lighter in relation to the head. It was probably 
slower-footed than the modern forms. 

Anterior Arch and Limbs. — The scapula varies somewhat in the series of 
specimens available, but is in general hardly to be distinguished from that of 
Canis lupus or C. occidentalis, excepting in its larger size. The region of the 
infraspinous fossa is commonly very wide, and the area for attachment of the 
teres major is distinctly marked. 

The humerus is a massive bone compared with that commonly seen in the 
Canidae. The deltoid ridge and the tuberosities are usually very strongly de- 
veloped. The ulna and radius are both heavy elements. The ulna is not dis- 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 237 

tinguished from that of the timber wolves by any sharply-marked characters. 
The head of the radius is, in general, relatively thick anteroposteriorly in cor- 
respondence to the anteroposterior thickness of the massive distal end of the 
humerus. On the distal end of the radius there is commonly only a faint groove 
for the tendon of the extensor ossis metacarpi pollicis, whereas in many modern 
wolves and coyotes this groove and the tubercle above it are clearly 
marked. The small size of the groove possibly indicates weakness of the extensor 
muscle of the thumb. A similarly shallow groove has been noted on one radius 
of a specimen of G. occidentalis. 

Study of a large series of specimens of C. dirus indicates that the feet 
are not relatively as heavy as in the large timber wolf, G. pambasileus. The 
mesopodial elements of the average specimens are of about the same size 
as those of a specimen of G. pambasileus in which the skull is somewhat smaller 
than in the average of G. dints. Compared with the same skeleton of C. pam- 
basileus, the average specimen of G. dims has absolutely smaller metapodial 
and phalangeal elements. 

In the mantis of G. dirus, excepting minor differences, -the form of the 
elements generally resembles that of the modern wolves quite closely. Among 
the carpals, the scapholunar seems commonly to be distinguished by the shorter 
transverse diameter of the medial facet of the distal side upon which the 
trapezoid articulates. This would seem to indicate a smaller transverse diam- 
eter of the trapezoid and a corresponding narrowing of the head of metacarpal 
two. It has not, however, been noted that either of these elements is markedly 
narrowed. The metacarpals in some cases have the shaft relatively wide 
anteroposteriorly in G. dirus. In metacarpal four the anterior medial facet at 
the proximal end is developed as a relatively long downwardly extended surface 
marked by a gentle elevation at about the middle of its length. In G. pam- 
basileus this facet is shorter vertically, and the protuberance is much more 
prominent. In some of the other canid forms, as in the domestic dog and in the 
coyote, this facet does not reach as far down on the shaft. 

So far as can be determined, the pollex is not larger, and is probably on the 
average smaller, than in the timber wolves. The finding of skeletons in which 
the elements are associated is very uncommon at Rancho La Brea, and it has 
been especially difficult to make certain of the association of the smaller ele- 
ments of the limbs. The suggestion that the pollex was small is based on the 
fact that the average of the specimens of metacarpal one in the collection is 
relatively small, this evidence being supported by that of the small groove for 
the tendon of the extensor ossis metacarpi pollicis, and the small, short facet 
for the trapezoid. 

The average of the phalangeal elements of the anterior limb in the collec- 
tion is smaller than in a large specimen of G. pambasileus. The terminal phal- 



238 



MEMOIRS OP THE UNIVERSITY OP CALIFORNIA. 



anges have stout, gently curved claw-cores, and are extended inf eriorly as strong 
subungular processes. 

Posterior Arch and Limbs. — The pelvic arch is closely similar to that of 
the timber wolves. In the posterior extremities the femur (fig. 18) varies 
considerably in the weight of the shaft, being in some cases heavier than in 
the timber wolves. The tibia is commonly very massive. In the femur the 
greater trochanter is always large in adult animals, and may project upward 
to a point almost level with the proximal side of the head. The smaller troch- 





18 



20 



K 



Fig. 18. Cants dims Leidy. Femur, posterior view, no. 10540. X V-. Rancho La Brea Beds. 
Pigs. 1!) to 21. Cam's dints Leidy. Tibia, no. 194S2. X '-. Fig. 19. "anterior side; fig. 20, medial view; fig. 21, 
superior view. Rancho La Brea Beds. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 



239 



anter and the spiral line below it are always well developed. The tubercle on 
the postero-lateral region of the shaft just above the popliteal surface is 
usually clearly marked, and is sometimes very large. Frequently the external 
epicondylic line passing through this tubercle is extended as a long, narrow 
sharp-edged ridge. 

The tibia (figs. 19, 20, and 21) is usually characterized by a very prominent 
cnemial crest which fades out interiorly in the lower portion of the upper 
third of this element. No peculiar characters are noted in the fibula. 

The astragalus has much the same form as in Ganis pambasileus, excepting 
in the vertical diameter of the neck. The distal face for articulation with 
the navicular is transversely elongated in G. dims to such an extent that the 
long diameter is about one-third greater than the short or vertical diameter. 
In C. pambasileus the neck and head are thicker, and the terminal face for 
articulation with the navicular is more nearly round. 

In the calcaneum the distal face for articulation with the cuboid tends to 
be a little narrower transversely and more nearly quadrate in form in G. 
pambasileus than in G. dims. The cuboid is a little narrower transversely 
iu proportion to its length, and the distal face for articulation with metatarsals 
four and five somewhat narrower in G. dims than in G. pambasileus. 
Corresponding to the form of the head of the astragalus the navicular is 
relatively narrow anteroposterior^ or vertically in G. dims. It is also noted 
that in this species the small posterior facet which meets the calcaneum is 
relatively larger than in G. pambasileus. The cuneiform elements are not 
materially different from those of the modern wolves. 

The metatarsals, like the metacarpals, are relatively somewhat shorter in 
the average specimen of C. dims than in G. pambasileus, 
though large specimens are present which exceed the largest 
measurements known in the latter form. The metatarsals 
of C. dims, even where shorter absolutely than in G. pam- 
basileus, are distinctly wider anteroposterior^ in the upper 
half of the shaft. Metatarsal I of G dims does not appear 
to show any noteworthy difference from that of the modern 
wolves. It is at least as large relatively as in G. pambasil- 1 
eus. Metatarsal II (fig. 23) is slightly narrower trans- 
versely at the proximal articular end than in G. pambasileus, 
and the angle on the median side of the shaft is less dis- 
tinctly marked than in that form. Metatarsal IV, in addi- 
tion to the greater anteroposterior width of the shaft, shows 
a sharper anteromedian angle in the middle third of the 
shaft. Metatarsal V (fig. 22) is very distinctly wider anter- 
oposterior ly in G. dims than in G. pambasileus and C. occi- 
dentalis. It is further characterized by the great promi- 




22 



23 



Figs. 22 and 23. Canis 
dints Leidy. Hetatar- 
sals, no. 19475, X ¥2. 
Fig. 22, right metatar- 
sal five, medial aspect; 
fig. 23, Tight metatar- 
sal two, lateral aspect. 
Rancho La Brea Beds. 



240 MEMOIRS OP THE UNIVERSITY OF CALIFORNIA. 

nenee of the tuberosity external to the proximal articular facet; in this char- 
acter it is, however, not distinctly different from C. occidentalis. 

As nearly as can be determined, relatively small size of the proximal pha- 
langes obtains in the hind feet, as in the anterior extremities. 

Measurements of Important Skeletal Parts 

The following measurements are taken from skeletal elements representing individuals 
of relatively large size. It is not certain that any two of the elements chosen represent the 
same individual. 

Atlas, greatest transverse diameter 120.5 mm. 

Axis, greatest anteroposterior diameter of neural spine 73.7 

Scapula, greatest height 201 

Pelvis, greatest length 228 

Humerus, greatest length 240 

Radius, greatest length 220 

Metacarpal four, greatest length 90.5 

Femur, greatest length 260 

Tibia, greatest length 237 

Metatarsal four, greatest length 102.3 

Comparison of Rancho La Brea Specimens with Previotjly Known 

Material 

Comparison with the Type Specimen. — In order to fix the systematic posi- 
tion of the California species satisfactorily it was considered of the utmost 
importance to make a comparison with Leidy's type specimen. Mr. TVitmer 
Stone of the Philadelphia Academy of Natural Sciences responded most cor- 
dially to a request for the loan of the original specimen, making it possible to 
compare the California collection with the type. 

As a result of the comparison of the type with a large number of the speci- 
mens from the asphalt deposits, there appears to be no essential difference 
between the Indiana and the California forms. Such differences as exist are 
hardly greater than the minor individual differences among California speci- 
mens evidently representing one species. 

In the type specimen (pi. 25, fig. 3) the form of M 1 , the most important 
tooth of the upper dentition, differs from that of the other wolves and resem- 
bles the California specimens in the extreme reduction of the hypocone. The 
form of this tooth is in general similar in the Indiana and California forms. 
If any difference is noticeable, it is in the slightly wider bases of the paracone 
and the metacone, and in the distinctly concave posterior border of the average 
California specimen. Numerous individuals are present in the California 
collection in which the outer pah' of tubercles does not show a broader base 
than in Leidy's type. The posterior border of most of the California specimens 
is generally rather sharply concave immediately behind the depression for 
the reception of the hypoconid of Mi. In the type the posterior border is very 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 241 

slightly concave at this point. In the type specimen the enamel has been 
almost entirely broken away from the outer and posterior sides of the metacone ; 
and as the extension of the cingulum around the postero-external angle of the 
tooth is largely removed, the border appears unnaturally straight. In a few 
California specimens the posterior border is almost as straight with the pos- 
tero-external portion of the cingulum present as it now appears in the type 
specimen with the enamel absent from this region. 

In the type specimen the cingulum of the anterior side of the paracone of 
M 1 appears relatively weak owing to the breaking away of the enamel imme- 
diately above the lower border of the cingulum. Judging from its sharp 
inferior ridge, the cingulum in this region was fully as strongly developed 
originally as in the Rancho La Brea specimens. 

In Leidy's description attention is called to the greater abruptness of the 
external portion of the basal ridge or cingulum than in the Recent wolves. In 
the type specimen the external portion of the cingulum of M 1 is marked by 
a fairly sharp ridge extending along the base of the paracone. This ridge is 
better developed than in most of the large modern wolves, and the inferior 
side of that portion of the cingulum bordering the posterior part of the base 
of the paracone forms a somewhat sharper ridge than in the average of the 
Rancho La Brea specimens. In some of the California specimens, however, 
the form and strength of the external cingulum correspond quite closely to what 
we see in the type. 

The form of M 2 in the type does not differ materially from that in the 
California specimens. The anterior end of the hypocone is extended around 
the inner and anterior side of the protocone, whereas in many of the Calif omia 
specimens it is interrupted on the antero-internal region of the tooth. There 
are, however, a number of individuals from Rancho La Brea in which pre- 
cisely the relations shown here are exhibited. In this tooth the metacone 
appears to approach the size of the paracone rather more closely than in most 
of the California specimens. The enamel being removed from both tubercles, 
it would be unsafe to accept the relative dimensions as certainly representing 
the true external form. Moreover, some of the Rancho La Brea specimens 
show practically the same relative dimensions of paracone and metacone as 
appear in M 2 of the type specimen. 

The superior carnassial of the type is badly broken, and shows nothing 
of the protocone blade. The antero-internal root does not extend far in toward 
the median line, indicating a small deuterocone as in the California specimens. 

The form of P 2 and P 3 of the type is closely similar to that of the California 
specimens. The extraordinarily strong internal cingula and the strong antero- 
internal ridges which Leidy mentions as occurring on these teeth in the type 
are noticed also in most of the Rancho La Brea specimens. 

The form of the maxillary bone of the type is not noticeably different from, 
that in the California species, excepting that the superior border of the infra- 



242 MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 

orbital foramen is situated a little lower than in the larger specimens from 
Rancho La Brea. 

Comparison with Cope's Texas Specimen. — The material which Cope de- 
scribed from the Texas Pleistocene includes very little on which to base a 
comparison. Fortunately the first upper molar (pi. 25, fig. 5), which is the 
most characteristic tooth in the whole dental series of C. dims, is represented. 
The dimensions of this tooth, as nearly as can be determined from Cope's 
figure of the specimen, are near to those of the type and to those of the Cali- 
fornia specimens. The general form of the tooth is clearly similar to that of 
both the type and the California material in the essential characters. The 
hypocone is greatly reduced, and the protocone is of moderate size, while the 
paracone and metacone are rather large. According to Cope's description 
and figure, the alveolus of the antero-internal root of the superior carnassal 
extends far forward so that its anterior border overlaps the posterior third of 
P 3 . Such a form in this root has not been exactly duplicated in the California 
material, although it is suggested in one specimen. If it is normal it would 
probably constitute a valid character distinguishing the Texas form. It 
is conceivable that the form of root seen here is abnormal. It seems quite 
unnecessary unless the deuterocone were very largely developed, and since 
the hypocone of M 1 is much reduced and the crushing power weak it is not 
probable that the deuterocone of P 4 was unusually developed. 

Cope noted that the Texas specimen differed from Leidy's type in the fol- 
lowing particulars: P 3 distinctly longer and the external eingulum weaker: 
internal root of P 4 extended farther forward ; protocone of M 1 less conic than 
in the type and external eingulum of M 1 weaker. He suggested that the differ- 
ence in the external cingula might be due to age, and was inclined to consider 
the two individuals as representing the same species. 

In Cope's figure of the Texas specimen 10 there is a suggestion of a break 
at the postero-external angle of M 1 , which has carried away the eingulum 
at this point. If this is the case, the anteroposterior diameter of M 1 would be 
shortened, and the length of P 3 would appear relatively large. The character 
of the external eingulum is found to be quite variable in the Rancho La Brea 
series of specimens. The nature of the antero-internal root of P 4 has been 
referred to above by the writer. The slight difference in the degree of lateral 
compression of the protocone of M l is fully equalled among the individuals 
of the Rancho La Brea series. 

So far as evidence is available, there seems to be good reason for consider- 
ing the California and Texas forms as closely related, if not identical species. 

Comparison with Material from Sheridan Formation. — A specimen in the 
collections of the American Museum which was obtained in the Sheridan f orma- 
tion of Kansas evidently represents Canis dims. It consists of a lower jaw 

io Cope, E. D., Jour. Acad. Sc. Pliilad., ser. 2, vol. 9, pi. 21, fig. 15, 1S95. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 



243 



(no. 10391 Amer. Mus. Nat. Hist.) with well worn dentition. Measurements 
of the mandible and dentition correspond closely with those of average speci- 
mens of C. dims from Rancho La Brea, as is shown in the table of measure- 
ments on page 232. The mandible is exceptionally high and thick as in G. dims. 
The massive sectorial is indistinguishable in form from that of C. dims, as 
is also M 2 . P 4 is noticeably large compared with the corresponding tooth in 
C. occidentalis, and in this respect resembles C. dims. The Sheridan Beds are 
presumed to represent an early phase of the Pleistocene. 

Comparison with Material from the Valley of Mexico. — Of considerable 
interest in connection with a study of the distribution of the American Pleisto- 
cene fauna is the occurrence in the Valley of Mexico of remains indistinguish- 
able from the specimens of Canis dims from Rancho La Brea. 




Fig. 24. Canis dims Leidy. Posterior portion of a cranium, 
Freudenberg. 



X %. Tequixquiac, Mexico. Adapted from 



The recent investigations of Freudenberg 17 have shown the existence in 
Mexico of a varied mammalian fauna, which resembles in many respects the 
Pleistocene fauna of the California region. Included in the assemblage are 
several specimens of a large wolf. Photographs and a cast of the best pre- 
served specimen 13 which Dr. Preudenberg very kindly placed at the writer's 
disposal represent an animal which does not appear to be specifically dis- 
tinguishable from the Rancho La Brea form of C. dims. No remains of the 
dentition of the Mexican type seem to have been obtained, so that a fully satis- 
factory determination of the characters is not possible, but the close relation- 
ship of the forms is evident from such comparisons as can be made. Especially 
noticeable are the similarity in size, in the form of the occipital region, and iu 
the nature of the overhanging inion region (fig. 24). The last character seems 
to be practically diagnostic of the C. dims group. In the large number of speci- 

n Freudenberg, W., Geol., u. Palae. Abh., N.F., Bd. 9, Heft 3, 1910. Die Siiugetkierfauna des Pliocans und 
Postpliocans von Mexico. 

is See Freudenberg, ibid., Taf. 6, figs. 2, 3, 4. 



244 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 



mens available in the collections from Rancho La Brea the individuals unques- 
tionably included in the species referred to C. citrus show considerable range 
of size and form. The Mexican species falls well within the range of variation 
of the Rancho La Brea specimens. The marked angle in the forehead of the 
Mexican specimen, formed by the sharp downward slope of the f ronto-maxillary 
region above the orbit, is matched b}^ the contour of the skull in several speci- 
mens representing old individuals from Rancho La Brea. 

Comparison tvith Previously Described Calif ornian Material. — The speci- 
mens already described from California, comprising the mandible from Liver- 
more Valley 19 and the fragment of a lower jaw from Tulare County 20 , may both 
presumably be included within the limits of Canis dims. The jaw from Liver- 




Fig. 25. Canis dims Leidy. Right ramus of mandible, X %. Livermore Valley, California. Adapted from 
Leidy. 

more Valley (fig. 25), as shown by the table of measurements below, is closely 
similar in dimensions to specimens from Rancho La Brea. The specimen from 
Tulare County (fig. 26) seemed to show a somewhat heavier jaw than the 




Fig. 2(3. Canis dims Leidy. A portion of the left ramus of the mandible, X %. Oil Springs, Oil Canon, 
Tulare County, California. 



i» Leidy, J., Geol. Surv. Terrs., vol. i, Foss. Verts, p. 230, 1873. 

20 Morriam, J. C, Univ. Calif. Publ. Bull. Dept. Geol.. vol. 3, no. 14. p. 288, 1903. 



MBRRIAM: THE FAUNA OF RANCHO LA BREA. 245 

individuals from Rancho La Brea. This may have been due in part to crush- 
ing, as the specimen was much broken. The measurement of the canine is much 
smaller than that in the larger individuals from Rancho La Brea, but was 
evidently taken somewhat higher up on the cone of the tooth. Considering 
the amount of variation known in the specimens at Rancho La Brea, there 
seems to be good reason for including both the Livennore Valley and Tulare 
County specimens in the group of Cants dims. 

Comparative Measurements 



No. 10856, larg 
specimen, Rancl 
La Brea 


No. 10834, medii 
specimen, Ranch 
La Brea 


CD 

"3 
> 

cd 

o a 
g 2 

P'3 
> o 

H-l CD 


*-* 

O 

o 

o 

© £ 

~ © 

fcr 1 to 


% 73 

= - 

7. T. 
'■£. -rH 

z Z 


219ap 


204 


200. mm. 




198 


48 


44 


44.1 




43 


39.7 


37 


37.5 


42 


39.5ap 


148 


137.7 


137.5 




133ap 


119.5 


113 


112.5 




UOap 


17.5 


16.5 


17.5 


13 




35.7 


34.5 


34.1 


35 


33.7 



Length of lower jaw from condyle 
to anterior side of canine 

Depth of lower jaw at condyle .... 

Depth of lower jaw at M x 

Length from posterior side of M 3 
to anterior side of canine 

Length of inferior molar and pre- 
molar series 

Anteroposterior diameter of in- 
ferior canine 

M x , anteroposterior diameter 

ap approximate. 

A fragment of a large wolf jaw (no. 5018) obtained in the Pleistocene 
deposits of Potter Creek Cave, Shasta County, California 21 shows some of the 
characters of Canis dims, as is indicated in the table of measurements below. 
The inferior carnassial is larger than that of the Recent wolves, but the heel 
is much narrower than in the typical G. dims, and the inferior premolars are 
relatively very small. The only specimen known is so fragmentary that final 
judgment as to its affinities should probably be withheld until better material 
can be obtained, but there does not seem to be sufficient evidence available to 
warrant definite separation of the Potter Creek form from the group of G. 
occidentalis. 



2i Univ. Calif. Publ. Am. Arch. Ethn., vol. 2, no. 1, p. 17, 1904. 



246 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 

Comparative Measurements 



M x , greatest anteroposterior diameter 

M 1; transverse diameter of heel 

P 4 , greatest anteroposterior diameter 

P 4 , greatest transverse diameter 

P ? , greatest anterposterior diameter 
P„, greatest anteroposterior diameter 

ap approximate. 

(«) No. 115995 U. S. Nat. Mus. 

(6) No. 8321. Calif. Mus. Vert. Zool. 



> 

US 


C3 


e3 


o 


1- 




M 
o 


M 


pq 


Oj 


CO C3 


t- a 


to h H 
o ^ % 


10 „ ij 

S 3 O 


CM J 

c. 3 O 


No. 5 
Pottei 
specin 


No. 11 
C. dir 
Ranch 


No. 1' 
C. dir 
Ranch 


ap33 mm. 


35.7 


32 


9.5 


13.5 


11.4 


15.3 


20 


19.8 


7.3 


10.5 


9.3 


13.2 


16.7 


15.8 


12.3 


15.4 


15 



'S c 


s £ 

3 -*-* 

CO B 

>— ' o 

. a< 

OP3 


32 


29.7 




10 


16 


16.5 


8.5 


7.7 





14.6 




12.5 



CANIS MILLBRI, N. Sp. 

Text figures 27 to 31 

Type specimen, a skull with lower jaw, no. 11257, University of California 
Collections in Vertebrate Palaeontology. Found four feet below the surface 
in the asphalt deposits at Rancho La Brea. The species is named in honor of 
Dr. L. H. Miller, whose assistance in furthering the investigations at Rancho 
La Brea has been greatly appreciated by the writer. 

Skull and dentition (figs. 27, 28, 29, 30, and 31) intermediate between 
Canis occidentalis and G. dims. Characters differing from those of G. dims as 
follows: Skull much smaller. Nasals narrower posteriorly. Nasal processes 
of f rontals broader and extending much farther forward. Fronto-f acial region 
relatively flat, and postorbital processes of frontals small. Sagittal crest low. 
Overhang of inion relatively small. Sharp median ridge of occiput wanting. 
Posterior palatine foramina relatively far forward. Lower jaw slightly nar- 
rower anteriorly than in typical C. dims. Teeth relatively large, P 4 and Mi 
especially heavy. M 1 (fig. 31) with unusually large hypocone, which extends 
around the antero-internal region of the protocone, and is connected with a 
low shelf of the cingulum on the anterior side of the tooth. P? with a distinct 
posterior cusp. Mi relatively large. 

The skull of Canis milleri differs markedly from that of C. occidentalis in 
its greater width, especially in the palate, and in the much more massive den- 
tition. P 4 and M 1 are much larger in proportion to the size of the skull than 
in any of the true timber wolves. P 4 greatly exceeds the relative dimensions 
in typical G. occidentalis. In the general form and massiveness of P 4 , this 
species closely approaches C. dims, from which it is, however, clearly separ- 
ated by other characters as indicated above. The nearest approach to the 
assemblage of characters seen in G. milleri is found in C. occidentalis furlongi. 
From this form G. milleri is distinguished by greater width of skull, especially 
in the palatine region ; greater interorbital width of the frontal region ; shorter 
and wider nasals; longer nasal processes of the frontals; somewhat heavier 
upper carnassials; and wider inner lobes of the upper molars. 

The form seen in skull no. 11257 was at first considered as probably repre- 
senting a young female of Ganis dims, the peculiar assemblage of characters 
observed being in part such as might occur in young individuals of this 
sex. Careful analysis of the collection does not, however, support this view. 

[247] 



248 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 





Figs. 27 and 28. Canis millcri Merriam, J. C. Skull, no. 11257, X %. Fig. 27, lateral view; fig. 28, superior 
view. Eancho La Brea Beds. 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 



249 




29 




Figs. 29 and 30. Canis milleri Merriam, J. C. Skull, no. 11257, X %. Pig. 29, posterior view; fig. 30, inferior 
view. Eancho La Brea Beds. 



250 MEMOIRS OF THE UNIVERSITY OP CALIFORNIA. 

A study of between forty and fifty skulls of C. dims shows no other 
specimen exhibiting the characters seen here, though there are available well- 
preserved specimens representing all stages of development, from forms with 
milk dentition to very aged individuals with teeth nearly worn away. In the 
collection of specimens unquestionably referred to C. dims there is much 
variation in many of the characters, particularly in size, and there is reason 
to believe that among the adults a number of the lighter skulls with weaker 
muscle attachments represent females. It is, to say the least, highly improbable 
that skull no. 11257 represents the only female in this collection. That the 
characters of youth added to those of sex are competent to extend the range 
of form in G. dims so far as to reach the assemblage of characters seen in 
specimen no. 11257 seems also improbable. This skull is that of a young adult 
much beyond the stage of development of many C. dims specimens in which 
the typical characters of the species are strongly expressed. Moreover, the 
variation in form of the nasals and nasal processes of the frontals is just 
opposite to that which would be expected in youthful animals. In all young 
specimens the nasal processes of the frontals are exceptionally short, while in 

no. 11257 they are larger than in any other indi- 
viduals, including those of advanced age. 

The Miller wolf was an animal about as large 
as the modern timber wolf, but with a relatively 
shorter and heavier head. It is to be presumed that 
the living animal differed very noticeably from the 

Pig. 31. Cams millcri Merriam. J. C. °_ _ • " 

Superior camassiai and superior dire wolf in size and in general contour of the body. 

molars, occlusal view, no. 11257, _ „ , . -, r . 1n ,„ , .,, . . 

natural size. Rancho La Brea Of the Miller wolf as yet we know with certainty 
Beds- only the skull. The species was evidently a rela- 

tively rare form at Rancho La Brea. but may have been much more common 
in other regions at the time this deposit was forming. 

For measurements of Cards milleri see table of comparative measurements 
of skull, p. 227. and dentition, p. 232. 




CANIS OCCIDENTALS FURLONGI Merriam, J. C. 
Text figures 32a to 33 & 

Canis occidentalis furlongi Merriam, J. C. Univ. Calif. Publ. Bull. Dept. GeoL, vol. 5, p. 393, 
1910. 

There are in the collections from Rancho La Brea several fragmentary 
spechnens representing a wolf species near Canis occidentalis. One of these 
specimens was made the type of a new subspecies in the publication cited 
above. The type material of the Rancho La Brea form when compared with 
typical C. occidentalis exhibits a tendency to relative narrowness of the nose, 
the superior carnassial tends to be relatively massive, and the second upper 
molar seems relatively narrow anteroposteriorly. 

The Rancho La Brea specimens originally referred to C. occidentalis fur- 





32a 



32b 



32c 

Figs. 32a to 32c. Canis occidentalis furlongi Merriam, J. C. No. 11283. Fig. 32a, portion of the skull with 
dentition, inferior aspect, X %; fig. 326, M 1 and M 2 , occlusal view, natural size; fig. 32c, portion of lower 
jaw with molars, X x /». Rancho La Brea Beds. 



longi differ from C. dims particularly in the form of M 1 (figs. o2a and 326). 
This tooth is relatively wide transversely, the inner lobe is relatively narrow 
anteroposteriorly, and the hypocone is relatively large. The hypocone has 
approximately the size and form seen in average specimens of wolves in the 
C. occidentalis group, and as in that species the anterior end of the hypocone 
ridge swings forward around the anterior side of the protocone instead of 
being interrupted as in G. dirus. 

[251] 



252 



MEMOIRS OP THE UNIVERSITY OP CALIFORNIA. 



In a fragment of a mandible (fig. 32c) accompanying the upper jaw specimen 
most clearly resembling C. occidentalis, the carnassial shows a metaconid even 
weaker than that of G. dims, while the entoconid is slightly larger than in 
average G. dims specimens. On M 2 the protoconid seems slightly smaller and 
the heel region relatively larger than in G. dims, though the metaconid is small 
compared with the protoconid. The proportions of the talonid region with 
reference to the trigonid are much as in C. occidentalis. 

In one of the small specimens (no. 10733) of the G. occidentalis type from 
Raneho La Brea the hypocone of M 1 is rather small, though larger than in the 
typical. G. dims, and the anterior extension of the hypocone ridge around the 
anterior side of the protocone is barely interrupted. M 2 is in this specimen 
of the narrow form with small metacone and hypocone. P 3 differs from the 
corresponding tooth of G. dims in being very narrow instead of wide pos- 





Figs. 33a and 336. Canis occidentalis furtonqi Merriam, ,T. C. Xo. 19792, X V->. Fig. 33a, skull superior view; 
fig. 336, skull inferior view. Eaneho La Brea Beds. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 253 

teriorly, and in the almost entire absence of a posteiior basal tubercle behind 
the posterior cusp. The portion of the palatine region present seems to narrow 
anteriorly. This specimen varies in some respects toward 0. dims, but the 
size, general form, and especially the proportions of M 1 indicate that it belongs 
with the group of individuals referred to C. occidentalis furlongi rather than 
to G. dims or to G. milleri. 

A small wolf skull, no. 19792, of the G. occidentalis type, obtained at Rancho 
La Brea by Dr. L. H. Miller, resembles G. occidentalis furlongi in the char- 
acters of the upper molars, and is referred to that form. This specimen (figs. 
33a and 33&) differs widely from G. dirus, and in most characters in which it 
differs from G. dirus, it resembles G. occidentalis. The skull is much smaller 
than in C. dirus, and is also relatively narrower. The nasal bones are long and 
narrow. The postorbital processes of the frontals are small, the sagittal crest 
is low, and the inion does not show the extraordinary overhang so characteristic 
of G. dirus. The posterior narial opening is narrow at the anterior end, instead 
of flaring as is commonly seen in G. dirus. The posterior palatine foramina 
are situated relatively far forward as in G. occidentalis. The molars are of the 
G. occidentalis form. P 4 is, however, much heavier than in the modern wolves 
and resembles closely P 4 of C. milleri. In the northern wolves of the G. pamba- 
sileus type the upper carnassial may be quite massive, but falls considerably 
below the stage of development seen in this form. 

Although there is a noticeable variation in the size of the teeth in the Rancho 
La Brea specimens referred to C. occidentalis furlongi, this material seems to 
represent a single form which differs from the modern gray wolves at least to 
the extent of subspecific variation. 

This wolf type was evidently relatively rare compared with C. dirus, and 
was much less common than the coyotes in this particular region. From what 
is known of the distribution of wolves of the G. occidentalis type it is doubtful 
whether they have ever been relatively abundant in the region of southwestern 
United States, as C. dirus was certainly the dominant species through a consid- 
erable part of Pleistocene time, and the coyote group has apparently been the 
most abundantly represented canid type since the disappearance of G. dirus. 



254 



MEMOIRS OP THE UNIVERSITY OP CALIFORNIA. 



Measurements of Dentition and Skull 



Length, posterior side of superior canine 

to posterior side of M 2 

Length, anterior side of P 4 to posterior 

side M 2 

P 1 , anteroposterior diameter 

M 1 . anteroposterior diameter along outer 

border 

M 1 , greatest transverse diameter 

M\ anteroposterior diameter along outer 

border 

M-', greatest transverse diameter 

Width, from outer side of alveolus of MP 

to median line 

Width, from outer side of alveolus of P 1 

to median line = 

Mj, anteroposterior diameter 

Mj, thickness measured across protoconid 
ML,, anteroposterior diameter 

(a) No. 10S34 

(b) No. 11257 

(c) No. 1308. U. S. Nat. Mus. 



pq 

CO d 

i—! O 


pq 

OS iJ 

G> ° 


pq 

CO c$ 

£ 2 


2 o 

.So 

t a 

. ci 

OP3 


:?cq 
■a a 

£ ° 

11 

OPh 


a 

o t> 
OPh 


81.5 mm. 


87 




115.5 


86 


91 


44.7 


48.5 




54 


47.7 


46 


23.6 


26.8 


24 


30.7 


28.2 


25.5 


16 


17.2 


15.5 


18.S 


17 


17.4 


19.3 


21.5 


18 


23 


20.7 


19.4 


8 


7.9 


8.2 


9.2 


9 


8.5 


10.8 


13.5 


10.5 


14.4 


12.9 


12.3 


39 


39.6 


37.7 


46.5 


41.S 


38.3 


18 


20.2 




27.5 


22.5 


19.5 


27 






34.5 


32 


28.3 


11.5 






13.6 


13.5 


11 


11.5 






13.3 


12.4 


11.7 



CANIS OCHROPUS ORCUTTI Merriam, J. C. 

Text figures 34 to 40 
Canis orcutti Merriam, J. C, Univ. Calif. Publ. Bull. Dept. Geol. vol. 5, p. 391. 1910. 

Type specimen, no. 10842, University of California Collections in Vertebrate 
Palaeontology. Prom the asphalt beds of Rancho La Brea, California. 

The specimens referred to this form are next in number to those repre- 
senting Canis dirus in the beds at Rancho La Brea, but are relatively rare 
compared with remains of that species. Though several factors may have 
tended to keep the number of coyotes entangled in the asphalt down to a 
percentage of the whole population somewhat less than in the case of the 
great wolves, it is probable that the smallness of the number of coyotes recov- 
ered is due mainly to an absolutely much smaller representation of these forms 
in this region during the time of accumulation of the asphalt beds. 

It is hardly to be presumed that coyotes would avoid the flat land bordering 
the hills to such an extent as to reduce tbe percentage of individuals entangled 
much below that in the case of G. dirus, though this might be true of the timber 
wolves. The principal factors which may have contributed to keep down the 
percentage of coyotes captured by the asphalt seem to be, nature of the lure 
attracting wolves, mode of hunting, and possible difference in intelligent recog- 
nition of the danger encountered. Most of the carnivores engulfed in the tar 
have been captured in one of three wa}^s: by accidental crossing of soft tar 
pools, by the lure of water pools in association with tar springs, and by the lure 
of entangled animals which might serve as food. Accidents under the first 
two heads would occur with about the same frequency in the two groups of 
wolves unless in one of them a grade of intelligence was developed which 
enabled the individuals to obtain a relatively better knowledge of danger signs. 
Whether the coyote was the more intelligent animal is not easily determined. 
It did, however, possess a relatively larger, though absolutely smaller r brain. 
It is not improbable that its sight, hearing, and smell were more acute than 
in the great wolf. If this be true, there is reason to suspect that the coyote 
would more readily perceive and avoid a danger not unusual in this region. 

The third factor, lure of animals, is the only one of the three which seems 
to have significance worth more than passing mention iu this connection. 
Judging from such evidence as we have, it seems probable that the great 
wolves were powerful enough to prey upon animals of considerable size, that 
they were so constructed as to make the tearing apart of large animals fairly 

[255] 



256 



MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 



easy work, and that they were numerous enough to make hunting in packs 
a natural method of attack. The coyotes evidently preyed upon small mammals 
and birds, and hunted alone or in small groups. In all of the asphalt collections 
brought together thus far the number of individuals representing the larger 
mammals has been unexpectedly great compared with that of the smaller forms, 
and the number of birds which would naturally serve as food for coyotes is 
also small. It may therefore be true that the lure for large wolves was excep- 
tionally good. It is to be noted, however, that even compared with a small 
number of individuals representing the smaller mammals the number of coyotes 
is small. It seems therefore possible to explain the number of coyotes present 
either on the supposition that, owing to much superior intelligence, out of a 
large number relatively few succumbed to accidental encountering of the tar, 
or to the attraction of living bait, or on the theory that the number of coyotes 
in the region was absolutely very much smaller than that of the great wolves. 
The latter view seems to give the principal reason for the small representation. 





35 

Figs. 34 and 35. Cants ochropus orvntti Merriam, J. C. Skull, no. 10842, X ¥>■ Fig. 34, lateral view; fig. 35, 
posterior view. Raneho La Brea Beds. 

This subspecies is closely related to Canis ochropus now living in southern 
California. The skulls of Canis ochropus orcutti average somewhat larger than 
in the living C. ochropus, and are noticeably broader across the palate and zygo- 
matic arches (figs. 31, 35, and 10). The mandible is considerably higher, par- 



MERRIAM: THE FAUNA OP RANCHO LA BREA. 



257 



ticularly below the molars, and is also thicker transversely than in the living 
form of this region. The dimensions of the teeth do not vary greatly from 
the living species excepting in the thickness of both the upper and lower car- 
nassials, which are heavier in the fossil form (figs. 36a and 36b). M 1 tends 
also to be somewhat heavier than in the typical G. ocliropus, and in this respect 




36a 



36b 



Pigs. 36a and 366. Cants ocliropus orcutti Merriam, J. C. Fig. 36a, superior carnassial and molars, occlusal 
view, no. 19791, natural size; fig. 366, inferior cheek-tooth dentition, occlusal view, no. 10842, X %. Kaneho 
La Brea Beds. 



more nearly approaches the typical G. latrans. In Mi the metaeonid seems 
to be slightly less prominent medially than in the typical G. ocliropus, possibly 
owing to the greater thickness of the trigonid blade in the fossil form. 

The form from Rancho La Brea differs from the typical G. latrans and 
resembles the type of G. ocliropus in the relatively narrow anteroposterior 
diameter of M 1 . M 2 is sometimes smaller than in either the typical latrans or 
the typical ocliropus form. 

A skeleton of this species which has been assembled from parts of separate 
individuals shows little difference from that of the living G. ocliropus of this 
region. It is evident that the coyote of Rancho La Brea was a slender-legged 
creature, and was swift-footed like its living relative. 






Fig. 37. Canis ochropus orcutti Merriam, J. C. Inferior view of atlas, no. 10S42, X %. Baneho La Brea Beds 
Fig. 38. Canis ochropus orcutti Merriam, J. C. Lateral view of axis, no. 10S42, X %. Baneho La Brea Beds 
Fig. 39. Canis ochropus orcutti Merriam, J. C. Metatarsus, X %• 



Baneho La Brea Beds. 



258 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA. 

In the collection of canid forms from Eancho La Brea there are a number 
of fragmentary specimens representing parts of the cranium of small coyotes 
which are hardly to be distinguished from the corresponding regions in the 
skull of typical G. ochropus. It is not impossible that two forms, typical G. 
ochropus and C. ochropus orcutti, were present. It is also possible that all of 
the forms of Eancho La Brea should be included in one variety which should 
go under the name of the living ochropus. From such material as is available 
the writer is, however, inclined to believe that the Eancho La Brea form differs 
somewhat from the living type ; and if all of the individuals are to be included 
in one subspecies, the use of the name C. ochropus orcutti for the group more 
truthfully represents the facts than would designation as typical C. ochropus. 
With a much larger series of specimens available it is possible that a further 
separation of the coyotes would be possible. 



MERRIAM: THE FAUNA OF RANCHO LA BREA. 



259 



Measurements of Skull and Dentition 



Length from anterior end of premaxillaries 

to posterior side of occipital condyles.... 

Width across zygomatic arches 

Width between outer sides of tritocones 

of P 4 

Least width between superior borders of 

orbits 

Width between postorbital processes of 

frontals 

Length, posterior side of superior canine to 

posterior side of M 2 

Length, anterior side of P 4 to posterior 

side of M 2 

P 3 , anteroposterior diameter 

P\ anteroposterior diameter 

P 4 , thickness across protocone 

M 1 , anteroposterior diameter measured 

along outer border 

M 1 , greatest transverse diameter 

M 2 , anteroposterior diameter measured 

along outer border 

M 2 , greatest transverse diameter 



<3 pj g AD<> 

^ g m ^ % (3 
o %£ g ",§ 

. a; c; . £3 & 

o bs>a osa 

. b rt . s ■« 

197.5mm. 188.5 
108 

65 

38 

55 

80.5 

37.3 

13.3 
21.2 

8.5 

13.3 

16 

7.3 

10.5 



— £> 

192 
104 



179.5 

106.8 



57 58.9 

35.4 29.6 

53.5 44 

80 72.6 

38.7 38.7 

13.2 13.2 



/. -_ - 



166.1 
91 



56 
31.9 
38.2 
64.5 ap 
36.5 ap 



20.8 


21 


20 


7.5 


7.7 


7.8 


12.8 


14 




16 


16.2 




8 


7.5 




11.3 


11.5 





No. 11278 



Length, anterior end of left ramus of mandible to 
middle of posterior side of condyles 

Height of mandible below posterior side of P 2 

Height of mandible below posterior side of M ± 

Thickness of mandible below protoconid of M 1 

Length, posterior side inferior canine to posterior 
side of M 2 

P 3 , anteroposterior diameter 

P 3 , greatest transverse diameter 

M 1; anteroposterior diameter 

M 1; greatest transverse diameter of trigonid 

M„, anteroposterior diameter 

(a) No. 12264. 
(6) No. 10842. 

(c) No. 651, Univ. Calif. Mus. Vert. Zool. 

(d) No. 10993. 

(e) No. 12249. 

ap approximate. 



C. ocbro- C. lat- 
pus (c) rans (d) 



145.5 mm. 


149 


140 


17 


16.3 


17.4 


22,5 


19.4 


20.9 


11.8 


10 


9.6 


85 


85 


77.5 


11.7 


11.5 


12.5 


4.8 


4.5 


4.S 


22.9 


22.2 


22.5 


9.5 


8.1 


S.3 


9.8 


9.S 


9.S 



CANIS ANDERSONI Merriam, J. C. 
Text figures 41 and 42 

Canis andersoni Merriam, J. C, Univ. Calif. Publ. Bull. Dept. Geol., vol. 5, p. 393, 1910. 
Type specimen no. 12249, University of California Collections in Vertebrate 
Palaeontology. Prom the asphalt beds of Rancho La Brea, California. 

A single specimen in the collections from Rancho La Brea represents a 
short-headed, coyote-like wolf quite different from any form known to the 
writer. The skull (figs. 41 and 42) is about as broad as that of Canis ocliropus 
but is relatively very short, with a relatively short and broad muzzle. This 
difference is noticeable also in comparison with the typical C. latrans. Though 





42 



41 




40 



Fig. 40. Canis ochropus orculti Merriam, J. C. Superior view of skull, no. 12264, X %. Rancho La Brea 
Beds. 

Figs. 41 ami 42. Canis andersoni Merriam. J. C. Skull, no. 12249, X %. Fig. 41, superior view; fig. 42, in- 
ferior view. Rancho La Brea Beds. 



[260] 



. MERRIAM: THE FAUNA OP RANCTIO LA BREA. 261 

this specimen represents a young individual, the permanent dentition had been 
complete and there is no reason to believe that the form and proportions of 
the skull would have changed materially in later life. Of the dentition only the 
superior carnassials have been preserved. These teeth have approximately the 
size of those in C. ocliropus, but appear slightly thicker. A number of minor 
differences between this specimen and typical representatives of the known 
coyotes may have specific or subspecific value, but their estimation is not 
possible with any degree of satisfaction when only one specimen is available 
for comparison. 

It is to be hoped that other material representing this form may be ob- 
tained so that some conception of the outlines of the body may be possible. 



Measurements of Skull and Dentition 

No. 12249 

Length from anterior end of premaxillaries to posterior side of occipital condyles 166.1 mm. 

Width across zygomatic arches 91 

Width between outer sides of tritocones of P 4 - 56 

Least width between superior borders of orbits — 31.9 

Width between postorbital processes of frontals 38.2 

Length, posterior side of superior canine to posterior side of M 2 64.5 ap 

Length, anterior side of P 4 to posterior side of M 2 - 36.5 ap 

P 4 , anteroposterior diameter 20 

P 4 , thickness across protocone — 7.8 

ay approximate. 



UROCYON CALIFORNICUS Mearns 

Text figure 43 



A finely preserved skull (fig. 43), no. 12263, represents a form almost- 
identical with the existing Urocyon calif ornicus of southern California. As 

is indicated in the tahle of measurements 
below, the dimensions are very close to those 
of two Recent specimens from the San 
Jacinto region not far distant. It is inter- 
esting to note that this form has survived 
to the present day with less modification 
than the other canid types. 

Portions of the lower jaw and dentition 

Fig. 43. Vrocj, Su- n f m ., n pvirlpntlv rpr>rp<?PTltiTl<r this wipHps 

perior view of skull, no. 12263, X %. Eanclio OI d I()1 m e\ lUCUUN lepiChLULlilg Llllb hpLCieb 

La Brea Beds. are a i so known from Rancho La Brea. 




Measurements of Skull and Dentition 



Length of skull, anterior end of premaxillaries to 

posterior side occipital condyles 

Width across zygomatic arches 

Least width between superior borders of orbits 

Width of palate between inner borders of second 

upper molars 

Length of superior dental series from anterior side 

of canine alveolus to posterior side of M 2 

Length from anterior side of P 4 to posterior side 

of M- 

P\ anteroposterior diameter along outer border .... 
M 1 , anteroposterior diameter along outer border .... 

M 1 , greatest transverse diameter 

M'-, anteroposterior diameter along outer border .... 
M-. greatest transverse diameter 

a Univ. Calif. Mus. Vert. Zool. 



No. 12263 


No. 2324a 


No. 2316a 


Kancbo La Brea 


Keeent 


Recent 


119.3 mm. 


118.7 


122.8 


68.2 


64.4 


67.6 


23.8 


22.5 


26 


17 


15.8 


18.4 


52.2 


52.5 


53.2 


22.3 


22.5 


22.3 


9.9 


10 




7.5 


8.4 


7.7 


10.6 


11.4 


10.5 


5.4 


6 


5.6 


7.9 


8.2 


7.4 



Date of issue, Octobt r 25, 1912. 



[262] 



PLATE 24. 



EXPLANATION OF PLATE 24 

Pig. 1. Canis dints Leidy. Skeleton approximately one-seventh natural size. Rancho La 
Brea Beds. 

Fig. 2. Canis ochropus orcutti Merriam. J. C. Skeleton approximately one-seventh natural 

size. Rancho La Brea Beds. 



1264 J 



MEMOIRS UNIV. CALIF., VOL. 1, NO. 2 



[MERRIAM] PLATE 24 





PLATE 25. 



EXPLANATION OF PLATE 25 
Canis dims Leidy 

Fig. 1. Skull, no. 10S34, lateral view. X %. Rancho La Brea Beds. 

Fig. 2. Skull, no. 19796, inferior view, X %• Rancho La Brea Beds. 

Pig. 3. M 1 and M 2 of the type specimen, natural size. 

Fig. 4. M 1 and M 2 of specimen no. 10856. natural size. Rancho La Brea Beds. 

Fig. 5. P 3 and M 1 of the Texas specimen described by Cope, natural size. 



[266] 



MEMOIRS UNIV. CALIF.. VOL. 1, NO. 2 



[MERRIAM] PLATE 25 









3 



4 



PLATE 26. 



EXPLANATION OF PLATE 26 

Canis dims Leidy. Eancho La Brea Beds 

Figures three-fifths natural size 

Fig. 1. Skull, no. 12266, superior view. 

Pig. 2. Skull of unusual relative breadth, no. 19796. superior view. 



[268] 



MEMOIRS UNIV. CALIF.. VOL. 1. NO. 2 



[MERRIAMj PLATE 2 





2 



PLATE 27. 



EXPLANATION OF PLATE 27 

Canis dints Leidy. Kancho La Brea Beds 

Figures three-fifths natural size 



Fig. 1. Scapula, outer side. no. 12963. 
Fig. 2. Pelvis, outer side. no. 19377. 



[270J 



MEMOIRS UNIV. CALIF., VOL. 1, NO. 2 



[MERRIAM] PLATE 27 





PLATE 28. 



EXPLANATION OF PLATE 28 

Ganis dims Leidy. Raneho La Brea Beds 

Figures approximately three-fifths natural size 



Fig. 1. Humerus, anterior view. 

Fig'. 2. Ulna, lateral view. 

Fig. 3. Radius, anterior view. 

Fig. 4. Femur, anterior view. 

Fig. 5. Tibia, anterior view. 



[272] 




- 









\) 




z 



to 




>K 




Ol 





> 



oo